Four investigators, each of who bring special expertise to the consortium, comprise the core of this proposal. The PI, Dr. Terry Van Dyke, has utilized mouse manipulation strategies to study cancer for 16 years. In the course of analyzing the tumor suppressors p53 and pRb, her lab has established several tumor models using transgenic and knock-out strategies. Dr. Eric Holland is a Neurosurgeon and newly established investigator. While a research associate in the Varmus lab, he developed a novel approach for short term somatic analysis of cancer genes in the mouse. He has made significant progress in defining growth factor roles in glial cell growth and migration. Dr David Louis is a Neuropathologist interested in the genetic lesions of human glioma. Most recently he mapped a new glioma specific tumor suppressor gene (TSG) on chromosome 19. Dr R Jude Samulski (UNC), a leader in genetic therapy, has developed viral vectors that show great potential for somatic gene delivery in vivo. The major experimental goal of this proposal is to develop mouse models that accurately reflect stages and classes of human glioma. Many strategies developed and utilized here can also be applied more broadly to other CNS cancer models. Dr Van Dyke's laboratory has already developed a choroid plexus (CP) tumor model that is histologically similar to the human disease. It describes the CP model and the approaches used for its derivation. Dr Holland's work will define the cellular responses to genetic changes observed in human glioma by somatic gene introduction into the mouse brain. The Van Dyke lab will utilize germline modification strategies to generate mouse strains with similar alterations. Close interaction between Van Dyke and Holland will facilitate rapid identification of genetic changes to be engineered into the germline. Preliminary results from these labs indicate significant progress toward producing glioma- like pathologies in the mouse. Project 4 will establish a glial- specific targeted inactivation of the pTEN gene, a common inactivation in human glioma. This mouse will be used in model derivation, and Dr Louis will assess the extent to which pathologies induced in the mouse resemble those observed in humans. Histological and genetic studies in human and mouse samples will be pursued. The Louis lab will complete their isolation of the glioma-specific TSG, and the Van Dyke lab will use gene-targeting strategies to assess its tumor suppressor function. Finally, we will examine a limited number of therapeutic strategies in existing and developing mouse models, including EGFR-targeted strategies by Dr Holland and anti- angiogenesis gene therapy by Drs. Samulski and Van Dyke.